Sure, you can tell your kids that it’s a reflection off the ocean, but this will really get them to shut the hell up.
Both colors are a result of light scattering, or when light hits something that can reflect it back in many directions. This is the difference between a nicely polished mirror and a pile of crushed mirror dust. One gives you a nice reflection, the second is a glittery mess. The glittery mess is a result of the light scattering.

A zillion tiny differences in the angle of incidence leads to scattering
The sky
The sky is blue due to a phenomena called Rayleigh scattering. Lord Rayleigh was an old school physicist who got a lot done, and gets many things named after him (his name also would have come up if anyone had cared to know how ink jet printers work, even though he lived and died ages before Canon Inkjets). Rayleigh scattering is an approximate description of how light gets scattered when it hits a small particle, say, nitrogen gas (N2). We’re talking small small, like so small that the wavelength of light (approx. 400-700 nm or 0.0000157 – 0.0000276 inches) matter, and matter a lot.
For unpolarized light (like from the sun) the intensity of light subject to Rayleigh scattering can be described by this equation, which I just stole from Wikipedia because it’s way cooler than the version I have in my notes:
I = I0*(1 + cos2(theta))/2R2 * (2pi/lambda)4*((n2-1)/(n2+2))*(d/2)
Where I0 is the original intensity, theta is the scattering, R is the distance to the particle, n is the refractive index, lambda is the wavelength in question, and d is the particle diameter. So for a particle of N2, I0, n and d are constant, cos2(theta) varies between 0 and 1, so that leaves us with R and lambda as variables of importance. I is inversely proportional to both, with I proportional to 1/R2 and 1/lambda4. The 1/R part tells you that the farther you are from the source (scattering particle) the weaker the light is. If I’m 1 unit away, I = AI0, where A is the rest of that shit. If I’m two units away, then I = (1/4)AI0. 3 units, I = (1/9)AI0 and so on. That’s pretty boring.
So let’s look at the other term: 1/lambda4, or 1/(lambda*lambda*lambda*lambda). That’s a lot of damn lambdas. The intensity really gets to feel every little change in lambda. If lambda = 1, then I = I0. But if lambda = 2, then I = (1/16)I0. Lambda should be measured in the same units as R and d (in fact, take a moment to notice that there is R2 and lambda4 dependence in the denominator, and d6 dependence in the numerator. If you measure R, lambda, and d in the same units, say meters, then you end up with meters^6/(meter^4*meters^2) = meters^6/meters^6 = unitless. The rest of the factors (cos(theta), n) are also unitless, because the whole mess is just a factor that reduces I0. If your units don’t cancel, then you turn I into I*m or something equally meaningless. The process of using units without numbers to make sure your answer will result in the proper form is called unit analysis, and is quite handy during exams when you’ve forgotten an equation).
Since I is proportional to lambda to the fourth power, any increase in lambda will result in a significant decrease in I. This means that longer wavelengths will scatter less than shorter, blue scatters more than red.
The snow
Snow is actually very simple compared to the sky. It’s got much more in common with our easily understood crushed mirror. Snow consists of snowflakes, which are crystals of ice with gaps of air. So a pile of snow is basically a skajillion bajillion umptillion fine icecubes with some space in between. We’ve already discussed, at great length, index of refraction, n. The ice has a larger n than the air, so the light will bend a decent amount as it passes through the many interfaces of air and ice.
So light hits a snow crystal, and some bounces off the top. But some goes through, and hits the other side, where it bounces off that, but some goes through and hits the next ice piece at a slightly different angle and bounces off of that, but some goes through and…you get the picture:

A single ray of light bounces through some snowflakes, simplified picture. Obviously the first reflection is the brightest, and the next ones get significantly weaker. But that’s OK, because there’s more than enough incident light for all of our light scattering needs
The end result is a glowing, sparkling, uniform white.
Recently